Research on the Evaluation of E-Commerce Cold Chain Food Consumption Based on Big Data

Abstract

In response to the problems of low efficiency and high cost of offline questionnaires, lack of keyword-consumer attitude correlation in online comment analysis, and little help for optimization solutions, an opinion analysis and improvement evaluation model based on sentiment analysis and Latent Dirichlet Allocation (LDA) document topic generation model is proposed. Using cold-chain food as the research object, a custom Python program was used to crawl the online consumer reviews about cold-chain food from Jingdong, a mainstream Chinese e-commerce company. A total of 70,134 reviews were obtained, including 65,535 valid reviews, which were analyzed by the LDA topic model and SnowNLP sentiment analysis to obtain the influencing factors and specific scores that affect consumer satisfaction. Then the importance ranking of the influencing factors was obtained by combining the word frequency and scores. Finally, based on the rankings and the reasons generated, suggestions are made for developing products and services for cold chain food. From theory and practice, it provides a reference basis for developing cold chain technology and consumer behavior research.